Created attachment 1215741 [details] fix x86 conditional The CpuInfoTest.SupportLevel testcase fails on s390x because wrong condition in src/gromacs/hardware/tests/cpuinfo.cpp is used, (IMHO) on unknown architecture the GMX_TARGET_X86 symbol is defined to zero, but I expect it's fully undefined on known non-x86 arches. [ 98%] Running all tests /usr/bin/ctest --output-on-failure Test project /builddir/build/BUILD/gromacs-2016/mpich Start 1: TestUtilsUnitTests 1/13 Test #1: TestUtilsUnitTests ............... Passed 0.05 sec Start 2: MdlibUnitTest 2/13 Test #2: MdlibUnitTest .................... Passed 0.01 sec Start 3: CommandLineUnitTests 3/13 Test #3: CommandLineUnitTests ............. Passed 0.02 sec Start 4: FFTUnitTests 4/13 Test #4: FFTUnitTests ..................... Passed 0.04 sec Start 5: HardwareUnitTests 5/13 Test #5: HardwareUnitTests ................***Failed 0.84 sec [==========] Running 5 tests from 2 test cases. [----------] Global test environment set-up. [----------] 1 test from CpuInfoTest [ RUN ] CpuInfoTest.SupportLevel /builddir/build/BUILD/gromacs-2016/src/gromacs/hardware/tests/cpuinfo.cpp:72: Failure Expected: (c.supportLevel()) >= (gmx::CpuInfo::SupportLevel::Features), actual: 4-byte object <00-00 00-01> vs 4-byte object <00-00 00-02> No CPU features could be detected. GROMACS might still work, but it will likely hurt your performance. Please mail gmx-developers so we can try to fix it. [ FAILED ] CpuInfoTest.SupportLevel (0 ms) [----------] 1 test from CpuInfoTest (0 ms total) [----------] 4 tests from HardwareTopologyTest [ RUN ] HardwareTopologyTest.Execute [ OK ] HardwareTopologyTest.Execute (831 ms) [ RUN ] HardwareTopologyTest.HwlocExecute [ OK ] HardwareTopologyTest.HwlocExecute (2 ms) [ RUN ] HardwareTopologyTest.ProcessorSelfconsistency [ OK ] HardwareTopologyTest.ProcessorSelfconsistency (2 ms) [ RUN ] HardwareTopologyTest.NumaCacheSelfconsistency [ OK ] HardwareTopologyTest.NumaCacheSelfconsistency (2 ms) [----------] 4 tests from HardwareTopologyTest (837 ms total) [----------] Global test environment tear-down [==========] 5 tests from 2 test cases ran. (837 ms total) [ PASSED ] 4 tests. [ FAILED ] 1 test, listed below: [ FAILED ] CpuInfoTest.SupportLevel 1 FAILED TEST Start 6: MathUnitTests 6/13 Test #6: MathUnitTests .................... Passed 0.01 sec Start 7: OnlineHelpUnitTests 7/13 Test #7: OnlineHelpUnitTests .............. Passed 0.01 sec Start 8: OptionsUnitTests 8/13 Test #8: OptionsUnitTests ................. Passed 0.01 sec Start 9: RandomUnitTests 9/13 Test #9: RandomUnitTests .................. Passed 0.03 sec Start 10: UtilityUnitTests 10/13 Test #10: UtilityUnitTests ................. Passed 0.02 sec Start 11: FileIOTests 11/13 Test #11: FileIOTests ...................... Passed 0.02 sec Start 12: PullTest 12/13 Test #12: PullTest ......................... Passed 0.01 sec Start 13: SimdUnitTests 13/13 Test #13: SimdUnitTests .................... Passed 0.02 sec 92% tests passed, 1 tests failed out of 13 Label Time Summary: GTest = 1.09 sec (13 tests) UnitTest = 1.09 sec (13 tests) Total Test time (real) = 1.10 sec The following tests FAILED: 5 - HardwareUnitTests (Failed) Errors while running CTest for full logs please see http://s390.koji.fedoraproject.org/koji/taskinfo?taskID=2388681 Version-Release number of selected component (if applicable): gromacs-2016-2.fc26
scratch builds succeed on s390x, ppc64le, x86_64 with the proposed fix applied http://s390.koji.fedoraproject.org/koji/taskinfo?taskID=2388828 http://koji.fedoraproject.org/koji/taskinfo?taskID=16258655 http://koji.fedoraproject.org/koji/taskinfo?taskID=16258616
reported upstream as https://gerrit.gromacs.org/#/c/6321/
seems fixed with gromacs-2016.2-1.fc26 - https://s390.koji.fedoraproject.org/koji/buildinfo?buildID=456201
This bug appears to have been reported against 'rawhide' during the Fedora 26 development cycle. Changing version to '26'.
This package has changed ownership in the Fedora Package Database. Reassigning to the new owner of this component.
This is fixed, according to comment #3.